[TxMt] Re: Huge files in Textmate

[Hans-Jörg Bibiko]

On 22.08.2008, at 15:19, Adam Strzelecki wrote:

>> This is my biggest (well, only) complaint about TextMate. I am
>> handling some sql dump files at the moment and TextMate runs out of
>> steam with files bigger than 5 MB. For these bigger files I use  
>> Oxygen
>> or Editix. I'd use TextMate for everything if it weren't for this
>> drawback.
>
> Do you happen to have "Check Spelling as You Type" on upon the
> opening? This may cause CPU drain issue with TM.
And renaming the file to *.txt helps sometimes.

But one could also do the following:
given a file 'test.sql' (150MB 670000 lines) and
a BASH script (only conceptually!)
- go to the file location
- mkdir split
- split -a 4 -l 10000 test.sql split/test.
- cd split
- for f in `find . -name 'test*'` ; do mv $f "$f".sql; done
- cd ..
- mate split
edit these files; make usage of Find/Replace in Project; or Ack in  
Project, Hypersearch, whatever
if the editing is finished
- cd split
- cat * > ../test_edit.sql

The script splits the huge file in files with 10000 lines each in a  
subfolder called 'split'.
Then open that subfolder via mate for editing. After editing join all  
files into test_edit.sql

Such a script could bound to APPLE+O. It checks if the file is greater  
than let's say 50MB, if so it splits it, if not it opens that file via  
mate.

The same strategy (see: man split) could also be used for huge XML  
files written in one line to insert a \n after a closing tag.

--Hans